Estimation of the covariance structure of heavy-tailed distributions
arXiv:1708.00502
Abstract
We propose and analyze a new estimator of the covariance matrix that admits strong theoretical guarantees under weak assumptions on the underlying distribution, such as existence of moments of only low order. While estimation of covariance matrices corresponding to sub-Gaussian distributions is well-understood, much less in known in the case of heavy-tailed data. As K. Balasubramanian and M. Yuan write, "data from real-world experiments oftentimes tend to be corrupted with outliers and/or exhibit heavy tails. In such cases, it is not clear that those covariance matrix estimators .. remain optimal" and "..what are the other possible strategies to deal with heavy tailed distributions warrant further studies." We make a step towards answering this question and prove tight deviation inequalities for the proposed estimator that depend only on the parameters controlling the "intrinsic dimension" associated to the covariance matrix (as opposed to the dimension of the ambient space); in particular, our results are applicable in the case of high-dimensional observations.
arXiv commentFixed some minor typos and inconsistencies
References in corpus (7)
- Nonlinear shrinkage estimation of large-dimensional covariance matrices
- Geometric median and robust estimation in Banach spaces
- A Shrinkage Principle for Heavy-Tailed Data: High-Dimensional Robust Low-Rank Matrix Recovery
- PAC-Bayesian bounds for the Gram matrix and least squares regression with a random design
- An Eigenvector Perturbation Bound and Its Application to Robust Covariance Estimation
- Robust Covariance and Scatter Matrix Estimation under Huber's Contamination Model
- PAC-Bayesian bounds for Principal Component Analysis in Hilbert spaces
Cited by in corpus (6)
- Distributed Estimation of Principal Eigenspaces
- Dimension-free PAC-Bayesian bounds for the estimation of the mean of a random vector
- Robust covariance estimation under norm equivalence
- Robust Modifications of U-statistics and Applications to Covariance Estimation Problems
- Affine Invariant Covariance Estimation for Heavy-Tailed Distributions
- On nearly assumption-free tests of nominal confidence interval coverage for causal parameters estimated by machine learning